of 12/12
ORIGINAL ARTICLE Qualitative and quantitative assessment of degeneration of cervical intervertebral discs and facet joints Joris Walraevens Baoge Liu Jos Vander Sloten Jan Goffin Received: 10 March 2008 / Revised: 14 August 2008 / Accepted: 17 October 2008 / Published online: 13 November 2008 Ó Springer-Verlag 2008 Abstract Degeneration of intervertebral discs and facet joints is one of the most frequently encountered spinal disorders. In order to describe and quantify degeneration and evaluate a possible relationship between degeneration and biomechanical parameters, e.g., the intervertebral range of motion and intradiscal pressure, a scoring system for degeneration is mandatory. However, few scoring systems for the assessment of degeneration of the cervical spine exist. Therefore, two separate objective scoring sys- tems to qualitatively and quantitatively assess the degree of cervical intervertebral disc and facet joint degeneration were developed and validated. The scoring system for cervical disc degeneration consists of three variables which are individually scored on neutral lateral radiographs: ‘‘height loss’’ (0–4 points), ‘‘anterior osteophytes’’ (0–3 points) and ‘‘endplate sclerosis’’ (0–2 points). The scoring system for facet joint degeneration consists of four vari- ables which are individually scored on neutral computed tomography scans: ‘‘hypertrophy’’ (0–2 points), ‘‘osteo- phytes’’ (0–1 point), ‘‘irregularity’’ on the articular surface (0–1 point) and ‘‘joint space narrowing’’ (0–1 point). Each variable contributes with varying importance to the overall degeneration score (max 9 points for the scoring system of cervical disc degeneration and max 5 points for facet joint degeneration). Degeneration of 20 discs and facet joints of 20 patients was blindly assessed by four raters: two neurosurgeons (one senior and one junior) and two radio- logists (one senior and one junior), firstly based on first subjective impression and secondly using the scoring sys- tems. Measurement errors and inter- and intra-rater agreement were determined. The measurement error of the scoring system for cervical disc degeneration was 11.1 versus 17.9% of the subjective impression results. This scoring system showed excellent intra-rater agreement (ICC = 0.86, 0.75–0.93) and excellent inter-rater agree- ment (ICC = 0.78, 0.64–0.88). Surgeons as well as radiologists and seniors as well as juniors obtained excel- lent inter- and intra-rater agreement. The measurement error of the scoring system for cervical facet joint degene- ration was 20.1 versus 24.2% of the subjective impression results. This scoring system showed good intra-rater agreement (ICC = 0.71, 0.42–0.89) and fair inter-rater agreement (ICC = 0.49, 0.26–0.74). Both scoring systems fulfilled the criteria for recommendation proposed by Kettler and Wilke. Our scoring systems can be reliable and objective tools for assessing cervical disc and facet joint degeneration. Moreover, the scoring system of cervical disc degeneration was shown to be experience- and disci- pline-independent. Keywords Disc and facet joint degeneration Á Scoring system Á Cervical spine Background Degeneration of intervertebral discs and facet joints is one of the most frequently encountered spinal disorders [20]. In order to describe and quantify degeneration and evaluate a possible relationship between degeneration and biome- chanical parameters, e.g., the intervertebral range of J. Walraevens (&) Á J. Vander Sloten Division of Biomechanics and Engineering Design, KULeuven, Celestijnenlaan 300C, PB 2419, 3001 Heverlee, Belgium e-mail: [email protected] B. Liu Á J. Goffin Division of Experimental Neurosurgery and Neuroanatomy, KULeuven, University Hospital Gasthuisberg, Leuven, Belgium 123 Eur Spine J (2009) 18:358–369 DOI 10.1007/s00586-008-0820-9

Qualitative and quantitative assessment

  • View
    214

  • Download
    0

Embed Size (px)

DESCRIPTION

Qualitative and quantitative assessment of degeneration of cervical intervertebral discs and facet joints

Text of Qualitative and quantitative assessment

  • ORIGINAL ARTICLE

    Qualitative and quantitative assessment of degenerationof cervical intervertebral discs and facet joints

    Joris Walraevens Baoge Liu Jos Vander Sloten Jan Goffin

    Received: 10 March 2008 / Revised: 14 August 2008 / Accepted: 17 October 2008 / Published online: 13 November 2008

    Springer-Verlag 2008

    Abstract Degeneration of intervertebral discs and facet

    joints is one of the most frequently encountered spinal

    disorders. In order to describe and quantify degeneration

    and evaluate a possible relationship between degeneration

    and biomechanical parameters, e.g., the intervertebral

    range of motion and intradiscal pressure, a scoring system

    for degeneration is mandatory. However, few scoring

    systems for the assessment of degeneration of the cervical

    spine exist. Therefore, two separate objective scoring sys-

    tems to qualitatively and quantitatively assess the degree of

    cervical intervertebral disc and facet joint degeneration

    were developed and validated. The scoring system for

    cervical disc degeneration consists of three variables which

    are individually scored on neutral lateral radiographs:

    height loss (04 points), anterior osteophytes (03

    points) and endplate sclerosis (02 points). The scoring

    system for facet joint degeneration consists of four vari-

    ables which are individually scored on neutral computed

    tomography scans: hypertrophy (02 points), osteo-

    phytes (01 point), irregularity on the articular surface

    (01 point) and joint space narrowing (01 point). Each

    variable contributes with varying importance to the overall

    degeneration score (max 9 points for the scoring system of

    cervical disc degeneration and max 5 points for facet joint

    degeneration). Degeneration of 20 discs and facet joints of

    20 patients was blindly assessed by four raters: two

    neurosurgeons (one senior and one junior) and two radio-

    logists (one senior and one junior), firstly based on first

    subjective impression and secondly using the scoring sys-

    tems. Measurement errors and inter- and intra-rater

    agreement were determined. The measurement error of the

    scoring system for cervical disc degeneration was 11.1

    versus 17.9% of the subjective impression results. This

    scoring system showed excellent intra-rater agreement

    (ICC = 0.86, 0.750.93) and excellent inter-rater agree-

    ment (ICC = 0.78, 0.640.88). Surgeons as well as

    radiologists and seniors as well as juniors obtained excel-

    lent inter- and intra-rater agreement. The measurement

    error of the scoring system for cervical facet joint degene-

    ration was 20.1 versus 24.2% of the subjective impression

    results. This scoring system showed good intra-rater

    agreement (ICC = 0.71, 0.420.89) and fair inter-rater

    agreement (ICC = 0.49, 0.260.74). Both scoring systems

    fulfilled the criteria for recommendation proposed by

    Kettler and Wilke. Our scoring systems can be reliable and

    objective tools for assessing cervical disc and facet joint

    degeneration. Moreover, the scoring system of cervical

    disc degeneration was shown to be experience- and disci-

    pline-independent.

    Keywords Disc and facet joint degeneration Scoring system Cervical spine

    Background

    Degeneration of intervertebral discs and facet joints is one

    of the most frequently encountered spinal disorders [20]. In

    order to describe and quantify degeneration and evaluate a

    possible relationship between degeneration and biome-

    chanical parameters, e.g., the intervertebral range of

    J. Walraevens (&) J. Vander SlotenDivision of Biomechanics and Engineering Design,

    KULeuven, Celestijnenlaan 300C, PB 2419,

    3001 Heverlee, Belgium

    e-mail: [email protected]

    B. Liu J. GoffinDivision of Experimental Neurosurgery and Neuroanatomy,

    KULeuven, University Hospital Gasthuisberg, Leuven, Belgium

    123

    Eur Spine J (2009) 18:358369

    DOI 10.1007/s00586-008-0820-9

  • motion, sagittal alignment and intradiscal pressure, a

    scoring system is mandatory. Moreover, a scoring system

    can be a helpful tool to investigate the possible correlation

    between intervertebral disc degeneration and facet joint

    degeneration or to assess the evolution of degeneration

    over time after an arthrodesis or after arthroplasty.

    However, as of date a limited number of scoring systems for

    degeneration of cervical intervertebral discs and facet joints

    based on radiographs have been developed [12]. Two scoring

    systems for cervical disc degeneration have been tested for

    reliability (Kellgren et al. [10] by Cote et al. [4] and Kettler

    et al. [11]). Only one scoring system for facet joint degenera-

    tion has been tested for reliability (Kellgren et al. [10]).

    In their review Kettler and Wilke observed a wide variety

    in design and terminology of the existing scoring systems

    [12]. One of the major drawbacks of the scoring systems of

    Kellgren for cervical disc and facet joint degeneration is the

    use of subjective, descriptive terms as moderate, severe

    to quantify degeneration. To ensure objectivity, Kettler et al.

    [11] developed a numerical radiographic scoring system for

    cervical intervertebral disc degeneration. In this scoring

    system three variables: height loss of intervertebral disc

    height, osteophyte formation and diffuse sclerosis have

    to be graded individually on a scale from 0 to 3. Based on the

    sum, the overall degree of disc degeneration is determined.

    Although Kettler and Wilke obtained substantial inter-rater

    agreement (j = 0.688), the scoring system has some draw-backs. Firstly, the scoring system is difficultly applicable in

    daily clinical practice because it is complex and time-con-

    suming. Secondly, the scoring system was developed based

    on the lateral radiographs of human cadaveric osteoliga-

    mentous spine specimens. It has not been tested in vivo.

    A reliable scoring system for the assessment of cervical

    facet joint degeneration does not exist up to date. Kellgren

    et al. [10] used lateral radiographs to score the degenera-

    tion of cervical facet joints. Cote et al. [4] found an

    unacceptable inter-rater agreement for this scoring system

    (ICC = 0.45). They claimed that one of the reasons for this

    poor agreement was that lateral radiographs often poorly

    visualize the facet joints. Computed tomography scans

    might improve the visibility and therefore early degenera-

    tive change might be better detected.

    The goal of this study is to establish and validate a

    quantitative scoring system for cervical intervertebral disc

    degeneration based on lateral radiographs and a scoring

    system for cervical facet joint degeneration based on

    computed tomography scans. The results of these scoring

    systems are compared with results based on first subjective

    impression. Moreover, as an application of both scoring

    systems, the spatial correlation between facet joint degene-

    ration and intervertebral disc degeneration is investigated.

    Materials and methods

    Scoring system for cervical disc degeneration based

    on lateral radiographs

    The scoring system for cervical disc degeneration consists

    of three variables with decreasing importance to the total

    degeneration score: height loss, anterior osteophytes,

    and endplate sclerosis. Each of these variables is indi-

    vidually scored. Next, the three variables are summed to

    the overall degree of disc degeneration (ranging from 0 to

    9; Table 1). Height loss is defined as the middle disc height

    with respect to a normal middle disc height at an adjacent

    level. Height loss is graded from 0 to 4. Middle disc height

    of the target level is assessed with respect to the middle

    Table 1 Scoring system ofcervical disc degeneration based

    on neutral lateral radiographs

    VB vertebral body,AP anteroposterior

    1. Height loss

    Middle disc height compared to normal middle

    disc height at an adjacent level

    0%

    B25%

    [25%B50%[50%B75%[75%

    0 points

    1 points

    2 points

    3 points

    4 points

    2. Anterior osteophytes with respect to

    the AP diameter of the corresponding VB

    No osteophytes

    B1/8 AP diameter

    [1/8B1/4 AP diameter[1/4 AP diameter

    0 points

    1 point

    2 points

    3 points

    3. Endplate sclerosis No sclerosis 0 points

    Detectable 1 point

    Definite 2 points

    Overall degree of disc degeneration = 1 ? 2 ? 3 0 points (no degeneration)

    13 points (mild degeneration)

    46 points (moderate degeneration)

    79 points (severe degeneration)

    Eur Spine J (2009) 18:358369 359

    123

  • disc height of a normal adjacent level (Fig. 1). The length

    of the anterior osteophytes is measured with respect to the

    anteroposterior diameter of the corresponding vertebral

    body, which is measured at the middle of the vertebral

    body (Fig. 2). Anterior osteophytes are scored from 0 to 3.

    When different scores are attributed to the cranial and

    caudal anterior corners of the target level, the highest score

    is chosen. For endplate sclerosis, a distinction between no

    apparent sclerosis, just detectable and definite sclerosis is

    made (score 02, Fig. 3).

    Scoring system for cervical facet joint degeneration

    based on computed tomography scans

    The scoring system for degeneration of the facet joints

    consists of four variables with varying importance to the total

    score: hypertrophy, osteophytes, irregularity of the

    articular surface, and joint space narrowing. These vari-

    ables are assessed on computed tomography scans. Similarly

    to the scoring system for cervical disc degeneration, each of

    these variables is individually scored. Next, the four

    variables are summed to obtain the overall degree of facet

    joint degeneration (ranging from 0 to 5; see Table 2).

    Hypertrophy is graded from 0 to 2. Zero points are given

    when there is no hypertrophy present; 1 point when

    hypertrophy is visible on one of the margins of the articular

    surface; and 2 points when it is present on all margins

    (Fig. 4). Osteophytes are graded 0 if no osteophytes are

    present and 1 if osteophytes are present (Fig. 5). Irregu-

    larity of the articular surface is scored 0 if the articular

    surface is smooth; and is scored 1 if the articular surface is

    irregular (Fig. 6). If the joint space of both facet joints is

    not narrowed, joint space narrowing is scored 0. In case of

    narrowing, it is scored 1 (Fig. 7). If a difference in

    degeneration score between the left and right facet joint is

    found, the highest of both scores is used.

    Experimental procedure

    In a retrospective study, neutral lateral radiographs and

    computed tomography scans of 20 patients, recently

    A

    B

    C

    D

    Fig. 1 Height loss is assessed on lateral radiographs. Middle discheight of the target level (CD) is compared to the middle disc heightof a normal adjacent level (AB). No height loss is scored as 0; a loss indisc height less than 25% receives 1 point; height loss between 25%

    and 50% receives 2 points; between 50% and 75% 3 points; and 4

    points are given when the height loss is more than 75%

    BA

    C D

    X Y

    P Q

    Fig. 2 Anterior osteophytes are assessed on lateral radiographs. Thelength of the anterior osteophytes (XY and PQ) is measured withrespect to the anteroposterior diameter of the vertebral body (AB andCD, respectively). When no anterior osteophyte is visible, a score of 0is attributed; an anterior osteophyte that is just detectable receives 1

    point; an anterior osteophyte which extends less than one-fourth of

    the anteroposterior diameter receives 2 points; when the anterior

    osteophytes extends more than one forth, 3 points are attributed. The

    highest of the cranial and caudal score is used as final score

    360 Eur Spine J (2009) 18:358369

    123

  • operated for cervical degenerative disc disease, were used

    for the assessment of intervertebral disc and facet joint

    degeneration. The name of each patient was removed and

    the clinical history remained unknown to prevent bias.

    Twenty intervertebral discs and facet joints of the oper-

    ated level were analyzed by four raters: two neurosurgeons

    (one senior and one junior) and two radiologists (one senior

    and one junior). None of the raters were previously con-

    nected to the study. Written instructions were provided to all

    raters before the assessment. No assistance was given during

    the assessment.

    Intervertebral disc degeneration and facet joint degen-

    eration were assessed three times: firstly based on

    subjective impression (SI; see Table 3), followed by a

    second time using the scoring systems (SS1). After

    1 month, the raters were asked to reassess all levels a third

    time using the scoring systems (SS2). Between all assess-

    ments, the order was randomized to prevent bias.

    Intervertebral disc degeneration was scored based on

    neutral lateral radiographs; facet joint degeneration for the

    operated level was assessed on computed tomography

    scans.

    Fig. 3 Endplate sclerosis is assessed on lateral radiographs. Zero points are attributed if no endplate sclerosis is present (a); 1 point is given ifsclerosis is just detectable (b); 2 points are given when sclerosis is definitively present at the cranial and/or the caudal endplate (c)

    Table 2 Scoring system ofcervical facet joint degeneration

    based on computed tomograph

    scans

    FJ facet joint

    1. Hypertrophy of FJ None 0 points

    On one of the margins

    of the articular surfaces

    1 point

    On all margins of the

    articular surfaces

    2 points

    2. Osteophytes on FJ None 0 points

    Yes 1 point

    3. Irregularity on articular surface Normal 0 points

    Irregular 1 point

    4. Joint space narrowing Normal 0 points

    Narrowed 1 point

    Overall degree of facet joint

    degeneration = 1 ? 2 ? 3 ? 4

    0 points (no degeneration)

    1 point (mild degeneration)

    23 points (moderate degeneration)

    45 points (severe degeneration)

    Fig. 4 Hypertrophy is assessed on transverse computed tomography scans. Zero points are given when no hypertrophy is present (a); 1 pointwhen hypertrophy is present on one of the margins of the articular surface (b); and 2 points when it is present on all margins (c)

    Eur Spine J (2009) 18:358369 361

    123

  • Statistical analysis

    The measurement error was estimated using within-subject

    standard deviations based on the SI and SS1 results. The

    measurement error is a measure for the variation in the

    scoring system [1, 2]. Ninety-five percent prediction limits

    can be calculated using the measurement error. The dif-

    ference between the observed value and the measured

    value is expected to be less than this value in 95% of the

    observations.

    Inter-rater agreement, i.e., the agreement between the

    ratings of all raters, and the intra-rater agreement, i.e., the

    agreement between the ratings of the same rater, were

    evaluated using a two-way random model of intra-class

    correlation coefficients (ICC), with measures of absolute

    agreement [18]. A single-measure intra-class correlation

    was selected to estimate the reliability of a single rating

    instead of a mean of several ratings. Inter-rater agreement

    was assessed based on the SI and SS1 results. Intra-rater

    agreement was calculated based on the SS1 and SS2

    results. Ninety-five percent confidence intervals (CI) were

    Table 3 Overall degree of degeneration of cervical disc and facetjoints based on first subjective impression

    Indication

    No degeneration

    Mild degeneration

    Moderate degeneration

    Severe degeneration

    Fig. 5 Osteophytes areassessed on transverse

    computed tomography scans.

    Zero points are given when no

    osteophytes on either of the

    facet joints are visible (a); 1point when osteophytes are

    present (b)

    Fig. 7 Joint space narrowing isassessed on computed

    tomography scans. Zero points

    are given when the joint space

    of either of the facet joints is not

    narrowed (a); 1 point when thespace is narrowed (b)

    Fig. 6 Irregularity of thearticular surface is assessed on

    transverse computed

    tomography scans. Zero points

    are given when the articular

    surface of either of the facet

    joints is smooth (a); 1 pointwhen the surface is irregular (b)

    362 Eur Spine J (2009) 18:358369

    123

  • constructed around each ICC [22]. Table 4 provides the

    convention that is used throughout the text.

    Linear correlations were investigated using Pearson r

    correlation coefficients. Data analysis was performed using

    Statistica 6.0.

    Results

    Disc degeneration

    Measurement error

    The scoring system shows an improved measurement error

    for the overall degree of disc degeneration the with regard to

    the SI result (11.1 vs. 17.9%; Table 5). The variable end-

    plate sclerosis has the largest measurement error (33.9%).

    Inter-rater agreement

    The inter-rater agreement for the overall degree of disc

    degeneration of the scoring system is excellent

    (ICC = 0.78, 0.640.88; Table 6). The ICC of the variable

    endplate sclerosis is poor (ICC = 0.31 vs. 0.73 for

    height loss and anterior osteophytes). The overall degree of

    disc degeneration shows a small improvement in ICC with

    respect to the SI results (ICC = 0.77, 0.590.89).

    Intra-rater agreement

    Excellent intra-rater agreement is observed for the overall

    degree of disc degeneration (ICC = 0.86, 0.750.93;

    Table 7). This observation holds for all variables indivi-

    dually, except for endplate sclerosis which has a good

    intra-rater reliability (ICC = 0.62, 0.400.80).

    Comparison between raters with different experience

    A comparison between experienced and inexperienced

    raters was made. Senior raters obtained better inter-rater

    agreement than junior raters for the variables height loss

    and endplate sclerosis (Fig. 8). Junior raters showed better

    inter-rater agreement for the SI results and anterior osteo-

    phytes. For all variables, except for endplate sclerosis,

    junior as well as senior raters obtained excellent intra-rater

    agreement (Fig. 9), with best results for the junior raters.

    Junior as well as senior raters obtained good inter-rater

    and excellent intra-rater agreement for the overall degree of

    disc degeneration (p [ 0.05).

    Comparison between raters of different disciplines

    A comparison between raters of different disciplines was

    made: one senior and one junior surgeon versus one senior

    Table 4 Convention for inter- and intra-rater agreement according toFleiss et al. [7] ICC: Intra-class correlation coefficient

    ICC Strength of agreement

    ICC B 0.40 Poor agreement

    0.40 \ ICC B 0.60 Fair agreement0.60 \ ICC B 0.75 Good agreement0.75 \ ICC Excellent agreement

    Table 5 Measurement error of cervical disc degeneration based on the assessment of 20 intervertebral discs [relative within-subject standarddeviation (WSSD) and 95% prediction limit (PL)]

    Scale WSSD (%) 95% PL

    Subjective impression of the overall degree of disc degeneration 17.87 1.0509

    Height loss 04 12.29 0.9635

    Anterior osteophytes 03 17.21 1.0121

    Endplate sclerosis 02 33.85 1.3269

    Overall degree of disc degeneration 09 11.11 1.9600

    Table 6 Inter-rater agreement between all raters based on the assessment of 20 intervertebral discs [Intra-class correlation coefficients (ICC) and95% confidence intervals (CI)]

    ICC 95% CI

    Subjective impression of the overall degree of disc degeneration 0.7650 0.59130.8874

    Height loss 0.7284 0.54030.8673

    Anterior osteophytes 0.7275 0.58130.8586

    Endplate sclerosis 0.3107 0.12730.5582

    Overall degree of disc degeneration 0.7759 0.64210.8871

    Eur Spine J (2009) 18:358369 363

    123

  • and one junior radiologist. The radiologists obtained better

    inter-rater agreement for the overall degree of disc

    degeneration (p \ 0.05). The inter-rater agreement of allvariables was better for the radiologists, except for anterior

    osteophytes. The intra-rater agreement of all variables was

    better for the radiologists, except for height loss.

    Surgeons as well as radiologists obtained excellent inter-

    rater and intra-rater agreement for the overall degree of

    disc degeneration (Figs. 8, 9).

    Facet joint degeneration

    Measurement error

    The overall degree of facet joint degeneration shows an

    improved measurement error in comparison to the SI result

    (20.1 vs. 24.2%; Table 8). Nevertheless, the measurement

    errors remain large for the overall degree of facet joint

    degeneration (20.1%) and all of the variables individually

    (39.343.6%).

    Inter-rater agreement

    Poor inter-rater agreement for the SI results (ICC = 0.35,

    0.160.59; Table 9) and fair inter-rater agreement of the

    overall degree of facet joint degeneration (ICC = 0.49,

    0.260.74) is obtained. Joint space narrowing was the

    variable with the highest inter-rater agreement

    (ICC = 0.40 compared to 0.17, 0.39 and 0.20 for hyper-

    trophy, osteophytes and irregularity).

    Intra-rater agreement

    Good intra-rater agreement was obtained for the overall

    degree of facet joint degeneration (ICC = 0.72, 0.42

    0.89). However, for the variables individually, with the

    exception of osteophytes, fair intra-rater agreement was

    found (Table 10).

    Comparison between raters with different experience

    A comparison between experienced and inexperienced

    raters was made. Senior raters obtained better inter-rater

    agreement than junior raters for the SI results and the

    overall degree of facet joint degeneration (Fig. 10).

    Table 7 Intra-rater agreement between two assessments (SS1 andSS2) of all raters based on the assessment of 20 intervertebral discs

    [Intra-class correlation coefficients (ICC) and 95% confidence inter-

    vals (CI)]

    ICC 95% CI

    Height loss 0.8039 0.60360.9137

    Anterior osteophytes 0.7869 0.58470.9034

    Endplate sclerosis 0.6156 0.39780.8010

    Overall degree of disc degeneration 0.8580 0.74610.9338

    0.0

    0.1

    0.2

    0.3

    0.4

    0.5

    0.6

    0.7

    0.8

    0.9

    1.0

    Subjectiveimpression

    Height loss Anteriorosteophytes

    Endplatesclerosis

    Overalldegeneration

    ICC

    * *

    Seniors Juniors Surgeons Radiologists 4 ratersFig. 8 Inter-rater agreementbetween senior versus junior

    raters and surgeons versus

    radiologists and between all

    raters based on the assessment

    of 20 intervertebral discs (Intra-

    class correlation coefficients

    (ICC); * p \ 0.05)

    364 Eur Spine J (2009) 18:358369

    123

  • Senior raters obtained excellent intra-rater agreement for

    the overall degree of facet joint degeneration, juniors

    obtained good agreement (p \ 0.05; Fig. 11).

    Comparison between raters of different disciplines

    A comparison between raters of different disciplines was

    made: one senior and one junior surgeon versus one

    senior and one junior radiologist. Surgeons obtained

    better inter-rater agreement than radiologists for the

    overall degree of facet joint degeneration and all of its

    variables, except for hypertrophy (Fig. 10). The surgeons

    obtained good inter-rater and excellent intra-rater agree-

    ment for the overall degree of facet joint degeneration

    and the radiologists fair and good agreement, respec-

    tively (Fig. 11).

    0.0

    0.1

    0.2

    0.3

    0.4

    0.5

    0.6

    0.7

    0.8

    0.9

    1.0

    Height loss Anterior osteophytes Endplate sclerosis Overall degeneration

    ICC

    *

    Seniors Juniors Surgeons Radiologists 4 ratersFig. 9 Intra-rater agreementbetween two assessments (SS1

    and SS2) of senior versus junior

    raters and surgeons versus

    radiologists and of all raters

    based on the assessment of 20

    intervertebral discs (Intra-class

    correlation coefficients (ICC); *

    p \ 0.05)

    Table 8 Measurement error ofcervical facet joint degeneration

    based on the assessment of 20

    facet joints [relative within-

    subject standard deviation

    (WSSD) and 95% prediction

    limit (PL)]

    Scale WSSD (%) 95% PL

    Subjective impression of the overall degree

    of facet joint degeneration

    24.25 1.4258

    Hypertrophy 02 30.37 1.1907

    Osteophytes 01 40.09 0.7857

    Irregularities 01 43.64 0.8554

    Joint space narrowing 01 39.34 0.7711

    Overall degree of facet joint degeneration 05 20.06 1.9658

    Table 9 Inter-rater agreementbetween all raters based on the

    assessment of 20 facet joints

    [Intra-class correlation

    coefficients (ICC) and 95%

    confidence intervals (CI)]

    ICC 95% CI

    Subjective impression of the overall degree

    of facet joint degeneration

    0.3494 0.16470.5902

    Hypertrophy 0.1708 0.00000.4759

    Osteophytes 0.3979 0.17070.6817

    Irregularities 0.2031 0.01280.5139

    Joint space narrowing 0.4007 0.17690.6837

    Overall degree of facet joint degeneration 0.4866 0.25890.7449

    Eur Spine J (2009) 18:358369 365

    123

  • Correlation between cervical disc and facet joint

    degeneration

    As shown in Table 11, a significant but weak correlation is

    observed between disc and facet joint degeneration based

    on the SI results (Pearson r: 0.33, p \ 0.05) as well asbased on total degeneration scores of disc and facet joint

    degeneration (Pearson r: 0.27, p \ 0.05).

    Discussion

    In this study separate scoring systems for cervical inter-

    vertebral disc and facet joint degeneration were proposed

    and tested for inter- and intra-rater agreement. Using these

    scoring systems, the spatial correlation between disc and

    facet joint degeneration was assessed.

    Only two scoring systems for the assessment of cervical

    intervertebral disc degeneration have been tested for inter-

    rater agreement up to date: Kellgren et al. [10] by Cote

    et al. [4] and Kettler et al. [11]. Both scoring systems

    obtained good inter-rater agreement (ICC = 0.71 and

    j = 0.688), compared to the excellent inter-rater agree-ment of our scoring system (ICC = 0.78).

    Our scoring system for cervical intervertebral disc

    degeneration has some similarities with the scoring system

    of Kettler et al. Three variables have to be graded indi-

    vidually on a numerical scale based on objective criteria;

    the sum of these scores assigns the overall degree of

    degeneration. Nevertheless, our scoring system is funda-

    mentally different. In contrast to the scoring system of

    Kettler et al., only middle disc height is used to calculate

    height loss. Moreover, only anterior osteophytes, but no

    posterior osteophytes, are assessed in our scoring system. A

    third difference is that the variables contribute to the

    overall degree of degeneration with variable importance.

    Height loss of the middle disc height has the highest

    importance (four points on a nine-point scale), followed by

    anterior osteophytes and endplate sclerosis (three and two

    points on a nine-point scale). This strategy was chosen

    because height loss is a straightforward and accurate

    indicator for disc degeneration. In contrast to posterior

    osteophytes, which are often not clearly visible on lateral

    radiographs due to the overlap with the lateral processes,

    anterior osteophytes are easily scored. Anterior osteophytes

    have however less clinical importance and therefore lower

    importance than for height loss was assigned in our scoring

    system. Endplate sclerosis contributes to disc degeneration,

    but the scoring of endplate sclerosis is very sensitive to the

    Table 10 Intra-rater agreement between two assessments (SS1 andSS2) of all raters based on the assessments of 20 facet joints [Intra-

    class correlation coefficients with 95% confidence intervals (CI)]

    ICC 95% CI

    Hypertrophy 0.5596 0.24000.8161

    Osteophytes 0.6568 0.38180.8576

    Irregularities 0.5443 0.21060.8040

    Joint space narrowing 0.5176 0.20670.7902

    Overall degree of facet joint degeneration 0.7167 0.42300.8894

    0.0

    0.1

    0.2

    0.3

    0.4

    0.5

    0.6

    0.7

    0.8

    0.9

    Subjectiveimpression

    Hypertrophy Osteophytes Irregularities Joint spacenarrowing

    Overalldegeneration

    ICC

    Seniors Juniors Surgeons Radiologists 4 ratersFig. 10 Inter-rater agreementbetween senior versus junior

    raters and surgeons versus

    radiologists and between all

    raters based on the assessment

    of 20 facet joints (Intra-class

    correlation coefficients (ICC); *

    p \ 0.05)

    366 Eur Spine J (2009) 18:358369

    123

  • quality of the radiographs and the proper alignment of the

    intervertebral disc. Therefore, the lowest importance was

    attributed to endplate sclerosis.

    These modifications did not lead to a lower inter-rater

    agreement. On the contrary, a stronger inter-rater agree-

    ment was observed. According to Kettler and Wilke, the

    inter-rater reliability of the scoring system fulfills their

    criterion for recommendation (ICC [ 0.60) [12].In addition to the inter-rater agreement, also intra-rater

    agreement of the scoring system for intervertebral disc

    degeneration was calculated. This value is a measure for

    the reproducibility of the scoring system. Excellent intra-

    rater agreement was observed for our scoring system.

    A drawback of this study is that no validation of the

    scoring system against a gold standard was performed.

    The excellent inter- and intra-rater agreement indicates

    that the scoring system is highly reliable and repeatable.

    However, such agreement does not eliminate the possibility

    of a systematic error (consistent over- or underestimation

    of the real degree of degeneration). Kettler and Wilke

    validated their scoring system based on lateral radiographs

    of human cadaveric osteoligamentous spine specimens, and

    used macroscopic slices of the respective cadaveric speci-

    mens to assess the real degree of degeneration. They

    found that the real degree of disc degeneration was

    underestimated in 64% of all discs [11]. However, the use

    of the cadaveric specimens might have influenced the

    results. The surrounding soft tissues that can decrease

    visibility of the intervertebral disc space are removed. And

    in contrast to in vivo measurements, a long exposure time

    can be used. This increases contrast on the lateral radio-

    graphs, providing a better visibility.

    Next to the cadaveric specimens, MRI might have been

    used as a comparative method to further validate our

    scoring system. Several scoring systems have been devel-

    oped to assess cervical intervertebral disc degeneration

    0.0

    0.1

    0.2

    0.3

    0.4

    0.5

    0.6

    0.7

    0.8

    0.9

    Hypertrophy Osteophytes Irregularities Joint spacenarrowing

    Overalldegeneration

    ICC

    *

    Seniors Juniors Surgeons Radiologists 4 ratersFig. 11 Intra-rater agreementbetween two assessments (SS1

    and SS2) of senior versus junior

    raters and surgeons versus

    radiologists and of all raters

    based on the assessment of 20

    facet joints (Intra-class

    correlation coefficients (ICC); *

    p \ 0.05)

    Table 11 Correlation table of Pearson r coefficients of intervertebral disc degeneration versus facet joint degeneration based on the subjectiveimpression (SI) results and the scoring systems. Significant correlations (p \ 0.05) are displayed in bold

    SI disc Height loss Anterior osteophytes Endplate sclerosis Overall

    degeneration (disc)

    SI facet joint 0.33 0.22 0.04 0.10 0.19

    Hypertrophy 0.36 0.26 0.08 0.20 0.27

    Osteophytes 0.45 0.38 0.10 0.21 0.35

    Irregularities 0.14 0.23 -0.09 -0.11 0.03

    Joint space narrowing -0.01 0.02 -0.07 0.07 0.01

    Overall degeneration (facet) 0.38 0.36 0.02 0.16 0.27

    Eur Spine J (2009) 18:358369 367

    123

  • based on MRI [5, 13, 16]. Miyazaki et al. [13] claimed that

    MRI is the most sensitive method for the clinical assess-

    ment of intervertebral disc pathology. They reported

    excellent intra-rater reliability and good to excellent inter-

    rater reliability for their scoring system. Similar to this

    study, Miyazaki et al. limited the validation of their scoring

    system to inter- and intra-rater reliability testing. No

    comparison against a gold standard, such as cadaveric

    specimens, was made. Four spinal surgeons acted as

    observers in their study; no information on their experience

    was given. In our study, both surgeons and radiologists,

    juniors as well as seniors acted as observers, illustrating the

    multi-experience and multi-discipline use of our scoring

    system. Christe et al. reported that both radiographs and

    MRI significantly, but weakly, correlated with histology

    (r = 0.33 and r = 0.49, p \ 0.05) in the detection ofpathologic lesions in the cervical spine [5]. They did,

    however, not report on the correlation between MRI and

    planar radiographs.

    As this is a retrospective study, no MRI was available

    for all patients. A comparison of our scoring system with

    MRI could therefore not be made.

    As this scoring system uses standard lateral radiographs,

    and as it requires uncomplicated input for the user, the

    scoring system can easily be used in daily clinical practice

    for the assessment of cervical disc degeneration.

    Senior and junior raters obtained good inter-rater

    agreement and excellent intra-rater agreement. Surgeons

    and radiologists obtained excellent inter- and intra-rater

    agreement. These results indicate that the scoring system

    can be reliably used by both experienced as well as inex-

    perienced raters from different disciplines.

    Only one scoring system for facet joint degeneration has

    previously been tested for reliability (Kellgren et al. [10]

    by Cote et al. [4]). Fair inter-rater agreement was found

    (ICC = 0.45). In contrast to Cote et al. [4], who believed

    that this level of agreement is not acceptable for rigorous

    outcomes research, Kettler and Wilke [12] noted that it

    fulfilled their criteria for recommendation (ICC [ 0.40).According to this criterion, also our scoring system for the

    assessment of facet joint degeneration (ICC = 0.49), can

    be recommended.

    Our scoring system assesses the presence of hypertro-

    phy, osteophytes, irregularities on the articular surface and

    joint space narrowing at the target level based on computed

    tomography scans.

    Similar to the scoring system of cervical disc degenera-

    tion, a drawback of this study is that the scoring system for

    cervical facet joint degeneration is not compared with a

    gold standard, such as cadaveric specimens, nor is it

    compared with an alternative method, such as MRI.

    As a clinical application, this scoring system is very

    useful when degeneration of one patient has to be assessed

    and compared at different time intervals, e.g., to investigate

    the influence of an arthrodesis or arthroplasty on the

    degeneration of the levels adjacent to the treated level.

    Moreover, this scoring system is applicable when degen-

    eration of different patients on a certain time point has to

    be compared. In these cases of relative comparison, a

    possible systematic error is canceled out.

    As an additional application for both scoring systems,

    the spatial correlation between intervertebral disc and facet

    joint degeneration has been investigated. In contrast to the

    lumbar spine [3, 6, 8, 9, 14, 15, 17, 19, 21, 23], this cor-

    relation has not been thoroughly investigated for the

    cervical spine. A weak but significant spatial correlation

    between cervical intervertebral disc degeneration and facet

    joint degeneration was observed. However, as this was not

    a follow-up study, the temporal correlation could not be

    identified. Therefore, the hypothesis that disc degenera-

    tion precedes facet joint osteoarthritis [3, 8, 23] cannot be

    confirmed, nor denied.

    Conclusion

    Our scoring system for cervical disc degeneration can be a

    reliable and objective tool. Moreover, this scoring system

    showed to be experience- and discipline-independent.

    Our scoring system for facet joint degeneration, which is

    based on computed tomography scans, is less reliable.

    Nevertheless, it fulfills the criteria for recommendation

    proposed by Kettler and Wilke.

    A weak spatial correlation between cervical interverte-

    bral disc and facet joint degeneration has been observed.

    Acknowledgments Medtronic Sofamor Danek Inc., USA, isacknowledged for the institutional research and education grant.

    References

    1. Bland JM, Altman DG (1996) Measurement error and correlation

    coefficients. BMJ 313(7048):4142

    2. Bland JM, Altman DG (1996) Measurement error proportional to

    the mean. BMJ 313:70497106

    3. Butler D, Trafimow JH, Andersson GB, McNeill TW, Huckman

    MS (1990) Discs degenerate before facets. Spine 15(2):111113.

    doi:10.1097/00007632-199002000-00012

    4. Cote P, Cassidy JD, Yong-Hing K, Sibley J, Loewy J (1997) Apo-

    physial joint degeneration, disc degeneration, and sagittal curve of

    the cervical spine. Can they be measured reliably on radiographs?

    Spine 22(8):859864. doi:10.1097/00007632-199704150-00007

    5. Christe A, Laubli R, Guzman R, Berlemann U, Moore RJ, Schroth

    G et al (2005) Degeneration of the cervical disc: histology com-

    pared with radiography and magnetic resonance imaging.

    Neuroradiology 47(10):721729. doi:10.1007/s00234-005-1412-6

    6. Derian TC (1994) Adjacent disc degeneration in patients with

    prior spinal fusion procedures. Spine 19(19):22442245. doi:

    10.1097/00007632-199410000-00024

    368 Eur Spine J (2009) 18:358369

    123

  • 7. Fleiss J (1986) Reliability of measurement. In: The design and

    analysis of clinical experiments. Wiley, New York, pp 132

    8. Fujiwara A, Lim TH, An HS, Tanaka N, Jeon CH, Andersson GB

    et al (2000) The effect of disc degeneration and facet joint osteo-

    arthritis on the segmental flexibility of the lumbar spine. Spine

    25(23):30363044. doi:10.1097/00007632-200012010-00011

    9. Gries NC, Berlemann U, Moore RJ, Vernon-Roberts B (2000)

    Early histologic changes in lower lumbar discs and facet joints

    and their correlation. Eur Spine J 9(1):2329. doi:10.1007/

    s005860050004

    10. Kellgren J, Jeffrey M, Ball J (1963) The epidemiology of chronic

    rheumatism, vol. II. Blackwell, Oxford, pp 1419

    11. Kettler A, Rohlmann F, Neidlinger-Wilke C, Werner K, Claes L,

    Wilke HJ (2006) Validity and interobserver agreement of a new

    radiographic grading system for intervertebral disc degeneration.

    Part II. Cervical spine. Eur Spine J 15(6):732741. doi:10.1007/

    s00586-005-1037-9

    12. Kettler A, Wilke HJ (2006) Review of existing grading systems

    for cervical or lumbar disc and facet joint degeneration. Eur

    Spine J 15(6):705718. doi:10.1007/s00586-005-0954-y

    13. Kolstad F, Myhr G, Kvistad KA, Nygaard OP, Leivseth G (2005)

    Degeneration and height of cervical discs classified from MRI

    compared with precise height measurements from radiographs.

    Eur J Radiol 55(3):415420. doi:10.1016/j.ejrad.2005.02.005

    14. Kumaresan S, Yoganandan N, Pintar FA, Maiman DJ, Goel VK

    (2001) Contribution of disc degeneration to osteophyte formation

    in the cervical spine: a biomechanical investigation. J Orthop Res

    19(5):977984. doi:10.1016/S0736-0266(01)00010-9

    15. Lewin T (1964) Osteoarthritis in lumbar synovial joints. a mor-

    phologic study. Acta Orthop Scand Suppl S73:112

    16. Miyazaki M, Hong SW, Yoon SH, Morishita Y, Wang JC (2008)

    Reliability of a magnetic resonance imaging-based grading sys-

    tem for cervical intervertebral disc degeneration. J Spinal Disord

    Tech 21(4):288292. doi:10.1097/BSD.0b013e31813c0e59

    17. Moore RJ, Crotti TN, Osti OL, Fraser RD, Vernon-Roberts B

    (1999) Osteoarthrosis of the facet joints resulting from anular rim

    lesions in sheep lumbar discs. Spine 24(6):519525. doi:

    10.1097/00007632-199903150-00003

    18. Shrout P, Fleiss J (1979) Intraclass correlations: uses in assessing

    rater reliability. Psychol Bull 86(2):420428. doi:10.1037/0033-

    2909.86.2.420

    19. Swanepoel MW, Adams LM, Smeathers JE (1995) Human

    lumbar apophyseal joint damage and intervertebral disc degen-

    eration. Ann Rheum Dis 54(3):182188. doi:10.1136/ard.

    54.3.182

    20. Szpalski M, Gunzburg R, Mayer M (2002) Spine arthroplasty: a

    historical review. Eur Spine J 11(S2):S65S84

    21. Tertti M, Paajanen H, Laato M, Aho H, Komu M, Kormano M

    (1991) Disc degeneration in magnetic resonance imaging a

    comparative biochemical, histologic, and radiologic study in

    cadaver spines. Spine 16(6):629634. doi:10.1097/00007632-

    199106000-00006

    22. Tian L (2005) On confidence intervals of a common intraclass

    correlation coefficient. Stat Med 24(21):33113318. doi:10.1002/

    sim.2145

    23. Vernon-Roberts B, Pirie CJ (1977) Degenerative changes in the

    intervertebral discs of the lumbar spine and their sequelae.

    Rheumatol Rehabil 16(1):1321

    Eur Spine J (2009) 18:358369 369

    123

    Qualitative and quantitative assessment of degenerationof cervical intervertebral discs and facet jointsAbstractBackgroundMaterials and methodsScoring system for cervical disc degeneration basedon lateral radiographsScoring system for cervical facet joint degeneration based on computed tomography scansExperimental procedureStatistical analysis

    ResultsDisc degenerationMeasurement errorInter-rater agreementIntra-rater agreementComparison between raters with different experienceComparison between raters of different disciplines

    Facet joint degenerationMeasurement errorInter-rater agreementIntra-rater agreementComparison between raters with different experienceComparison between raters of different disciplines

    Correlation between cervical disc and facet joint degeneration

    DiscussionConclusionAcknowledgmentsReferences

    /ColorImageDict > /JPEG2000ColorACSImageDict > /JPEG2000ColorImageDict > /AntiAliasGrayImages false /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 150 /GrayImageDepth -1 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict > /GrayImageDict > /JPEG2000GrayACSImageDict > /JPEG2000GrayImageDict > /AntiAliasMonoImages false /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 600 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict > /AllowPSXObjects false /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile (None) /PDFXOutputCondition () /PDFXRegistryName (http://www.color.org?) /PDFXTrapped /False

    /Description >>> setdistillerparams> setpagedevice